Artificial Intelligence: Less or Greater than Human Intelligence?

Nowadays computers become increasingly capable of solving numerous tasks that would otherwise require human intelligence. Deep learning in particular has demonstrated an impressive ability to recognize patterns in data like recognizing objects in images, spoken commands in audio, or understanding and translating text.

While current AI methods generalise well within the highly specific task they are trained for, they fail to infer anything about cause and effect and thus each task requires huge amounts of training data.

On the contrary a human does not need to experience jumping from the 5th floor to imagine the consequences. Unlike deep learning methods, a doctor can explain why particular physiological data may suggest a disease.

The question why things happen is still answered solely by human intelligence. Understanding cause and effect is key in creation of smarter AI.

Please see list of speakers and agenda here.